Feedback
freeTrialImage.bannerPity
freeTrialImage.upgradeUnlock
- ✓freeTrialImage.benefitHd
- ✓freeTrialImage.benefitWatermark
- ✓freeTrialImage.benefitUnlimited
MOSS-Image
Try MOSS-Image online: a 5B text-to-image model that reads long prompts and JSON layouts, placing every object, label and box exactly where you plan.
AI Tools
Explore AI tools
Free AI Image
Truly Free AI Image Generator
AI Image Editor
AI Photo Editor Online
Image Upscaler
Enhance image resolution effortlessly
AI Color Changer
Transform photo colors in seconds
AI Room Designer
Design your dream room with AI-powered 3D rendering
Skin Enhancer
Enhance skin texture naturally with AI
Face Editor
Enhance face texture naturally with AI
AI Clothes Changer
Why Teams Choose MOSS-Image for Controlled Renders
An open 5B family that turns written briefs and structured layouts into images you can position with precision.
- Layouts Built Into the ModelAssign each object and caption its own bounding box, and MOSS-Image follows your JSON rather than leaving composition to chance.
- Five-Billion-Parameter Core, 4K-Token PromptsLengthy briefs stay fully in context, and the same weights can be adapted to your own tasks and domains.
- Base and SFT CheckpointsBoth ship under Apache-2.0 with matching architecture; the SFT build keeps whiteboard lettering noticeably cleaner.
Running MOSS-Image in Three Steps
From a fresh Python environment to a layout-driven render in three moves.
Questions About MOSS-Image, Answered
Straight answers on prompts, checkpoints, hardware and licensing for MOSS-Image.
What exactly can I control with a JSON layout?
Each entry describes the scene, an object or a line of text, plus the box that fixes its position and size on the canvas.
What separates the Base and SFT checkpoints?
Same architecture, same sampling interface. SFT has been fine-tuned on curated data, which in the project demo yields tighter text grouping and a cleaner whiteboard.
How much text fits in one prompt?
Roughly 4K tokens in MOSS-Image — room enough for a full multi-element layout written as JSON.
Which hardware and software are required?
Python 3.10+ and an NVIDIA GPU with a matching PyTorch build. Low on VRAM? Switch to Sequential or shrink the image; Compile DiT stays off by default and is unavailable in that mode.
Does MOSS-Image work inside ComfyUI?
It does. Install the native plugin through ComfyUI Manager, then load the sample workflow; integration into the official repository is still underway.
Which license covers the weights and code?
Weights are published under Apache-2.0, while the pipeline code and third-party pieces keep their own terms. Cite the project repo in research work.
Put MOSS-Image to Work on Your Next Layout
Set up the Diffusers pipeline or the ComfyUI plugin, test a JSON layout prompt, or draft your idea with the built-in Prompt Helper before you render.